between Metallic Masses having different Temperatures. 443 



HOT COPPER. 



Cold Lead. Vibrates very well. 



Tin. The same. 



Zinc. Not at all; even where the temperature of the Copper is 



raised much above 212. 

 Iron. Vibrates very imperfectly. It appears, however, more active 



than Zinc. I have formerly observed a decisive vibration of 



Hot Copper upon Cold Iron (llth July). 

 Brass. No Vibration. 



HOT GOI,D. 



Cold Lead. Vibrates quite readily. Nearly as with Copper. 



Tin. Nearly as with Copper. 



Zinc. Ditto. 



Iron. No distinct vibration, but with considerable heat of the lamp 



approaches to it ; it is rather more disposed to vibrate than Zinc. 



By experiments analogous to these, we may see how the posi- 

 tion of any metal is fixed by a variety of tests, which afford 

 mutual confirmation. 



33. In the course of forming the classification of metals, I 

 was naturally led to compare it with Mr FARADAY'S hypothesis, 

 that the vibration between two metals depended on the diffe- 

 rence of their conducting powers for heat directly, and of their 

 expansion inversely. Finding considerable deviations from this 

 law, I was led to look for some simpler analogy. 



34. The first arrangement which presented a striking simi- 

 larity was that of the conducting powers of the metals for elec- 

 tricity. The further examination to which this remark led me, 

 pointed out another, and perhaps not less important, analogy, 

 which appeared not to have been before observed, namely, that 

 when the best data are collected, the order of conducting powers 

 of the metals for heat and for electricity is the same. I did not 

 adopt this conclusion till after a mature examination of existing 

 statements, and an extensive series of experiments upon the con- 



VOL. XII. PART II. 3 L 



